Traumatic brain injury (TBI) is a major national health problem. There are currently no effective therapies for TBI. Biochemical markers could provide confirmation of injury mechanism(s) and identify drug therapy targets. Temporal profiles of changes in biomarkers could guide timing of treatment. Biomarkers could provide clinical trial outcome measures obtainable more readily than neurological assessments. Biomarkers could also provide more acurate early diagnosis and prognosis, as well as provide end points for improving patient management in acute care and intensive care environments. There are no FDA approved biochemical markers for TBI. This proposal would initiate the first systematic identification and validation of a panel of biochemical markers for TBI detectable in human cerebrospinal fluid (CSF), a biological fluid routinely accessible following severe TBI. Specific Aim 1 will assess the diagnostic and prognostic utility of candidate biomarkers (breakdown products to alpha-spectrin-SBDPs; UCH-L1; MAP2a/b; myelin basic protein-MBP; S100B) by examining relationships between levels of specific biomarkers and injury magnitude (e.g. Glasgow Coma Scale-GCS), injury type (e.g. diffuse injury vs. mass lesion), occurrence of secondary insults (e.g. hypotension, elevated intracranial pressure) and outcome (e.g. neurological assessments and MRI). Sensitivity and specificity of markers, alone and in combination will be determined. Specific Aim 2 will assess the utility of SBDPs for assessment of oncotic and apoptotic celldeath by examining relationships between CSF levels of SBDPs and specific pathological events that occur after TBI. Specific Aim 3 will prospectively validate in a separate group of patients a diagnostic/prognostic panel of biomarkers selected on the basis of data from Specific Aim 1 by determining if the biomarker information will add diagnostic/prognostic information to the clinical data that is currently available and the information provided by individual biomarkers. [unreadable] [unreadable] [unreadable]